TCA9538: Can nRESET be tied directly to VCC without a pull-up resistor?

Hello,

While it is probably okay to connect the RESET pin straight to VCC, TI recommends using a pull up resistor in order to limit the inrush current into the device.

I can't make any guarantees that the device won't be harmed but based on how durable our chips are it will probably be ok.

Best,

Chris

Hi Chris. 

Thanks for getting back to me.  

Could you explain how the resistor limits inrush current?   The datasheet says that the maximum input current for the nRESET pin is 1uA, so even at 5V this pin should have 5MOhms of input resistance?

I must be missing something!  Something to do with the power supply ramping up changing the input resistance of nRESET?

Hello,

Let me know if I'm referring to the wrong thing but I think you are talking about this in the datasheet:

The electrical characteristics are not requirements of the device, it is simply how the device preforms with different test conditions. What this parameter talks about is the current going into or out of the pin when you only attach power to the device. You can think of it as the maximum leakage current from this pin.

For a pull up resistor we recommend anything from 1K-10K. This will limit the current going into the pin and will keep your pull up relatively strong.

Variac996 said:

Could you explain how the resistor limits inrush current?

This is one of the main purposes of resistors and it goes back to Ohms law. A resistance is equal to a voltage over a current. A resistor is essentially equal to how fast a specific potential of energy can move through it. If you know the voltage you are supplying to a pin, you can limit the current going into the pin by using a resistor. That's what Ohms law calculates.

The resistor takes the additional energy that it is limiting and turns it into heat.

If this answered your question don't forget to hit the button. If you have any more questions let me know.

Best,

Chris

Hi Chris.

Thanks for the reply.  I'm still a bit confused I'm afraid!

The highlighted section you correctly referred to says that at maximum 1uA of current will go in (or out) of the nRESET pin. Therefore, it must have an input resistance in the order of 6-mega-ohms, given that the maximum voltage that can be applied to the nRESET pin is 6V.  

Therefore, I cannot understand how having a 10kohm or 1kohm or any other pull-up resistor in the 10s of kohms range could make any appreciable difference to the in-rush current into the nRESET pin.  The effect of the resistance is neglectable compared to the existing input resistance.  (6Mohms+10kohms is basically still 6Mohms!)

Therefore, I'm clearly misunderstanding something about how this chip works.  The nRESET pin must be low impedance at some point during the operation of the TCA9538 for there to be any significant inrush current, presumably below Vcc = 1.65V? This operating condition is what I'm trying to understand as presumably the same effect could occur on other pins...

Sorry, I realize now I wasn't very clear in my post before.

To go back to my point from before, in the split second that you attach your power to your RESET pin, there will be a potential difference across your resistor. Eventually once both sides equal each other that potential difference will go away and that is when you see the input current you talk about above. But in the split second that there is a potential across your resistor that is when Ohms law comes into play and your resistor is limiting that current.

When you attach two different potentials together for a brief moment there will be a really high current as the higher potential charges up the lower potential. Everything, essentially, has some parasitic capacitance attached to it. So when you attached your power to your pin or you turn on the voltage, the parasitic capacitance of the RESET pin will have to charge up. What you see is a curve like this:

The derivative of this curve is the current going through that capacitor.

Your current is equal to V(t)/R. If you have no resistance then for a split second that inrush current is infinite. Now in reality, there also are resistors internally attached to the pin which is why you would probably be okay attaching the reset pin directly to VCC. However, we recommend adding an additional resistance to further limit that current.

Let me know if there is still any confusion.

Best,

Chris

Hi Chris. 

Thanks for the explanation, makes sense.

In general, for TI parts, is it safe to assume that control lines may be safely tied directly to VCC, if the datasheet does not specifically mention the requirement for a resistor?  

Hi Chris.

Thanks for checking and thanks for the help.

It's a bit ghastly if Ti datasheets don't explicitly mention the need for series R on inputs.  Might be worth raising this with Ti internally, given that things could blow up?  Bit challenging for designers to know what is safe, if the datasheets don't mention this.  I would expect the datasheets to cover all safety critical requirements.

Yes, that is correct, that is why we recommended to use a pull up resistor on the RESET pin.

TI is always going to make sure that any critical electrical requirements are stated in the absolute maximum ratings of the device. As long as you operate within those conditions you cannot do any permanent damage to the device. 

Otherwise, the recommended conditions are only there for the functionality of the device, so that the device can perform as you expect. If you operate outside the recommended conditions the device may not perform as you like but you will still not harm the device as long as you operate within the absolute maximum ratings.

I hope this clears up any worries about the information in TI datasheets.

Best,

Chris